The general aim of this project is to study the effects of immunotoxin induced transient T-cell depletion on the modulation of T cells from immunization responses towards tolerizing responses, and to use this knowledge for the experimental and clinical treatment of T cell driven autoimmune diseases, graft- versus-host disease, and the induction of tolerance to mismatched organ and cell transplants. We previously showed that a 2-3 day course of anti-rhesus CD3 immunotoxin constructed with CRM9, a binding site mutant of diphtheria toxin, depletes lymph node and blood T cells by 99% after a 48 hour lag period. This process was associated with a marked prolongation of survival of mismatched functioning rhesus kidney transplants and the induction of long term tolerance in about 50% of the cases without further immunosuppressive therapy. Rejections, when they occurred, were preceded by the appearance of anti-graft antibodies (done with S. Knechtle and J. Thomas). To circumvent this adverse T cell mediated B cell response that is apparently initiated during the delay in immunotoxin induced T cell killing, we have added a short course of agents that are believed to block antigen presentation such as deoxyspergualin (DSG). Long term graft survival has risen to 83% (done with J. Thomas). These grafts are tolerized as judged by the acceptance of donor skin grafts for 30 days and the rapid rejection of third party skin grafts. Congeneic pancreatic islet cell transplants in monkeys suffering from spontaneous insulin dependent diabetes have been shown to reverse diabetes as judged by returning non-fasting blood glucose and glycosylated hemoglobin values to normal in the absence of exogenous insulin. A short course of immunotoxin, methyl prednisolone and cyclosporine provide induction of stable operational tolerance to islets that has lasted over one year in one case. This is the first reported reversal of diabetes using a tolerizing protocol that does not depend on chronic immunosuppression, a process previously shown to compromise transplanted islet function (done with J. Thomas). Recombinant single chain anti-CD3 antibodies directed at both human and rhesus T cells have been developed. In eukaryote expression systems these are secreted as high affinity divalent dimers. These lack significant Fc receptor interactions and are free from the complications of cytokine release syndrome which occur with our present chemically coupled anti-CD3 immunotoxin. These recombinant antibodies form the basis of recombinant anti- CD3 immunotoxins now under development.